NomaDamas/CozyClay

Open source previs software in the browser: block a scene, pose characters, author camera moves and cuts, then take the same shots to an AI video model.

What it solves

CozyClay is a browser-based 3D staging studio that simplifies the process of blocking scenes, posing characters, and sequencing motion. It provides a local workspace where users can arrange set pieces, define character paths, and preview generated motion without needing a full-scale professional 3D suite.

How it works

Built with Three.js and React Three Fiber, the studio allows users to manipulate 3D primitives and characters using standard 3D editor gizmos. It integrates with a backend called Kimodo for AI-driven motion generation, where users can sequence "Prompt Blocks" on a timeline to create complex animations. Additionally, it implements the Model Context Protocol (MCP), allowing AI assistants like Claude to directly control the viewport, place characters, and generate motion using natural language commands.

Who it’s for

It is designed for filmmakers, animators, and creators who need to quickly prototype 3D scenes and sequence motion prompts for AI generation.

Highlights

  • AI-Driven Motion: Sequence multi-phase motion prompts on a timeline and generate animations via Kimodo.
  • MCP Integration: Direct the studio using an AI assistant to frame shots and place the cast via plain language.
  • ** uma 3D Staging Environment**: Full 3D editor capabilities including fly-cameras, grid snapping, and a comprehensive undo/redo history.
  • Low Friction Setup: Can be launched instantly via npx cozyclay with no complex dependency installation.

Related

  • Project
    OpenMOSS/MOSS-VL

    MOSS‑VL is an open‑weight 11 B‑parameter family of video‑language models that can understand and converse about live video streams in real time. It uses a cross‑attention architecture with timestamped frames and a 3‑D rotary embedding (XRoPE) to achieve low‑latency, interruptible dialogue, proactive silence, and dynamic correction. Three variants are released (Realtime, Instruct, Base), with quantized checkpoints for single‑GPU use, integration with FlashAttention‑3, SGLang, ms‑swift, and LoRA fine‑tuning. The project includes benchmarks showing state‑of‑the‑art streaming performance and provides scripts for both realtime and offline inference.

  • Project
    NoizAI/HelixWorld

    HelixWorld 1.0 is a forthcoming real‑time model that jointly generates video and spatial audio as a user moves a virtual camera. Trained on first‑person video/audio and game‑engine captures, it predicts the next frame and sound field causally, then distills the pipeline for interactive speeds. Code, weights, and a technical report are slated for release soon; the project will be Apache‑2.0 licensed.

  • Project
    theopenconversationkit/tock

    Tock is an open‑source conversational AI platform that provides an NLP stack, a visual studio for designing dialogue flows, a Kotlin‑based DSL (with Python, Node.js and REST bindings), ready‑made connectors to chat/voice channels, and Docker‑based deployment.

  • Project
    Cjbuilds/Codex-Orchestration

    Codex Orchestration is a Codex plug‑in that lets you attach other LLMs (Claude Fable 5, Opus 5, OpenRouter models, etc.) to a task and assign them specific roles—Planner, Advisor, Designer, Executor. Codex stays the top‑level orchestrator, iterating plans, reviewing them, optionally designing UX artefacts, and finally executing code. The plug‑in is installed via the Codex marketplace, configured with skill prompts inside a Codex chat, and supports role‑specific effort levels. It aims to improve plan quality, enable parallel work, and reduce premium‑model usage. MIT‑licensed.